Why Cooling Blowdown Is the Central Wastewater Problem in Pyongyang
Cooling-tower blowdown is the controlled purge from an evaporative cooling loop, drawn off to keep dissolved solids below scaling thresholds as pure water evaporates. Globally, evaporative losses account for roughly 80% of cooling makeup, with blowdown making up the remaining 20% (Valicor, citing Lawrence Berkeley National Laboratory's 2024 United States Data Center Energy Usage Report). U.S. data centers consumed approximately 17 billion gallons of water for cooling in 2023, and a 100 MW evaporative-cooled facility typically uses 400,000–550,000 gal/day, with summer peaks at Council Bluffs reaching 2.7 MGD (Valicor, citing Google's 2025 Environmental Report). That stream carries elevated total dissolved solids, hardness ions, silica, biocides, corrosion inhibitors, dispersants, and trace metals, per the June 2026 University of Georgia Cooperative Extension publication TP-121 by Saha, as cited by Commercial Water Lab.
For a Pyongyang site, blowdown, rather than sanitary or process wastewater, defines the entire treatment design. Chronic power rationing makes high-airflow closed-loop cooling harder to run reliably, so even a water-constrained operator is likely to keep some evaporative cooling online, and the resulting blowdown volume cannot be ignored. There are no public facility-level water-use disclosures for North Korean data centers, so design must rely on global benchmarks plus site-specific source-water testing. The 15–25% purge of recirculating cooling water that concentrates hardness, silica, treatment chemicals, and trace metals as pure water evaporates is the central design driver.
Pyongyang Site Constraints That Shape the Treatment Train
Pyongyang has cold winters and warm, humid summers, where seasonal peaks drive cooling load and blowdown volume, mirroring the Council Bluffs pattern that pushes to 2.7 MGD in summer (Valicor, citing Google 2025 Environmental Report). That seasonal swing means the treatment train must be sized for the wet-bulb peak, not the annual average, and the operator must plan for storage or throttling of the upstream feed.
Chronic grid instability and rationing favor low-energy, gravity-friendly treatment stages and limit the practicality of high-pressure, multi-stage RO skids that demand stable, high-kW input. The Taedong River and the municipal Pyongyang supply are the realistic feeds; turbidity and seasonal variability make robust pretreatment essential before any RO. Public permitting data for DPRK sewer and surface-water discharge is not published, so the design should default to maximum internal reuse and plan for zero or near-zero liquid discharge to reduce permitting exposure. Very limited specialty chemical supply chains and few trained operators mean the system must be fully automated, chemical-light, and tolerant of long intervals between servicing. These constraints—specifically low kW per cubic meter, low operator-hours per shift, and minimal imported consumables—rule out many standard Western blowdown designs.
The 2026 Treatment Train: From Side-Stream Filtration to RO Reuse
The 2026 baseline train for a Pyongyang data center is built in the order water actually flows, with each step anchored to an in-catalog equipment category.
| Step | Function | Equipment Category | Key Spec or Range |
|---|---|---|---|
| 1. Side-stream filtration on the cooling loop | Removes suspended solids; keeps blowdown quality manageable for downstream membranes | Multi-media filter | Capex $50,000–$200,000 for typical data center installations (Genesis Water Technologies) |
| 2. Ultrafiltration pretreatment | Protects RO from colloids, bacteria, and SDI excursions | Ultrafiltration pretreatment skid (0.03 µm PVDF hollow-fiber) | Turbidity acceptance up to 300 ppm with automatic backwash and air scour |
| 3. Antiscalant and chemical conditioning | Controls scaling chemistry ahead of RO | PLC-controlled chemical dosing skid | Closed-loop, low-consumable dosing tuned to feedwater analysis |
| 4. Reverse osmosis for cooling-tower makeup reuse | Rejects dissolved salts; permeate is reused as cooling makeup | Industrial RO unit | Industrial RO achieves up to 95% recovery (HydropureWater RO product spec); conventional BWRO on CTBD is limited to 75–80% recovery before scaling (IDE Tech) |
| 5. Disinfection of the reuse loop | Controls biological growth without chlorine-induced corrosion | Pipeline UV sterilizer | Chemical-free, on-line UV on RO permeate line |
| 6. Sludge handling at any precipitation step | Dewaters chemical solids to a handleable cake | Plate-and-frame filter press | Low-energy endpoint suited to a low-operator environment |
Step 1—side-stream filtration on the cooling loop, handled by a multi-media filter, removes suspended solids and keeps blowdown quality manageable for downstream membranes; the capex range cited in research is $50,000–$200,000 for typical data center installations (Genesis Water Technologies). Step 2—an ultrafiltration pretreatment skid using 0.03 µm PVDF hollow-fiber membranes protects the RO from colloids, bacteria, and SDI excursions; turbidity acceptance up to 300 ppm with automatic backwash and air scour is the relevant duty window for the seasonal Taedong feed. Step 3—a PLC-controlled chemical dosing skid conditions the RO feed with antiscalant and, if needed, a reducing agent to strip free chlorine, so the RO membranes see stable chemistry.
Step 4—an industrial RO unit handles the core recovery. The cited industrial RO product spec lists up to 95% recovery, and IDE Tech notes that on cooling-tower blowdown, conventional BWRO is limited to 75–80% recovery before scaling becomes unmanageable. Step 5—a pipeline UV sterilizer on the RO permeate line controls biological growth in the reuse loop without the corrosion risk that comes with free chlorine. Step 6—where the design includes any precipitation or lime softening, a plate-and-frame filter press is the practical, low-energy endpoint suited to a low-operator environment.
Where sewer discharge is restricted, route RO concentrate to a controlled-precipitation brine concentrator; IDE Tech reports about 95% recovery with permeate silica around 1 mg/L. Genesis Water Technologies cites $3–8M capex for full ZLD, which is likely over-scoped for a Pyongyang first build. A side-by-side comparison with the New York City data center blowdown guide and the Moscow data center blowdown guide shows that the DPRK case is differentiated by the operating envelope around the unit operations.
Blowdown Chemistry Targets and Reuse Decision Framework
Some jurisdictions enforce blowdown TDS below 1,500 mg/L before discharge is allowed (Genesis Water Technologies), and that number should be used as a worst-case benchmark when local DPRK limits are unknown. Set the reuse target next: RO permeate for cooling-tower makeup typically needs 10–50 mg/L TDS (Genesis Water Technologies), which both the cited 75–80% recovery BWRO and the ~95% recovery high-recovery systems can meet on appropriately pretreated CTBD. An industrial water softener ahead of the cooling loop and a high-efficiency sedimentation tank for any precipitation stage are the supporting equipment categories that buy chemistry headroom for the membranes.
| Site Condition | Decision | Driver |
|---|---|---|
| Sewer discharge permitted, stable 4 cycles of concentration achievable | Side-stream filtration + partial RO reuse; skip ZLD on the first build | Lowest capex that still meets the 10–50 mg/L permeate target |
| Discharge uncertain or restricted | Closed-loop reuse with brine concentration; accept higher capex to remove permitting risk | Reduces exposure where DPRK discharge rules are unpublished |
| Grid too unstable for reliable multi-stage RO | Downsize evaporative cooling, raise cycles of concentration, shift load to hybrid dry/wet cooling where economically possible | Matches equipment kW draw to available, unreliable grid |
| Blowdown TDS approaching the 1,500 mg/L ceiling | Add side-stream filtration and antiscalant dosing before pushing cycles higher | Avoids crossing the discharge threshold on raw blowdown |
The manufacturing water-use reduction guide covers the broader consumables picture for non-data-center sites.
2026 Cost, Risk, and Sizing Notes for a Pyongyang Project
Cost drivers from the cited research are direct discharge fees that can reach $5–$15 per 1,000 gallons, side-stream filtration capex of $50,000–$200,000, and ZLD capex of $3–8M (Genesis Water Technologies). DPRK projects will carry logistics and spares premiums on top of these numbers, so the buyer should request a site-adjusted quote, not a Western list price.
Sizing inputs to request from the EPC include facility IT load (MW), planned PUE and WUE, makeup water analysis (hardness, silica, TDS, iron, manganese, free chlorine), target cycles of concentration, and peak summer wet-bulb temperature. Compliance risk is high, as discharge to the Taedong River or to municipal sewer in Pyongyang is not documented in public DPRK regulation; treat any discharge path as requiring a pre-construction letter from the local utility and design for maximum reuse to reduce exposure. Schedule risk is elevated due to long lead times for RO membranes, control valves, and dosing pumps; the buyer should hold a 9–12 month critical-spares list in-country. The RO and UF membranes and filter elements catalog and the water treatment parts, valves, and media catalog are the realistic fast-replacement path.
Frequently Asked Questions
What is the minimum treatment train a Pyongyang data center actually needs in 2026?
Side-stream filtration on the cooling loop, UF pretreatment, an RO unit for cooling-tower makeup reuse, and UV disinfection on the permeate line, with a plate-and-frame filter press for any precipitation sludge. ZLD is only required where discharge is restricted or where local rules turn out to mirror the cited 1,500 mg/L TDS ceiling (Genesis Water Technologies).
How much should a 2026 buyer budget for the blowdown treatment train?
Use the cited ranges as a starting band: $50,000–$200,000 for side-stream filtration (Genesis Water Technologies) and $3–8M for full ZLD if it is required (Genesis Water Technologies). Request a site-adjusted quote rather than relying on a Western list price, because DPRK logistics and spares premiums apply, and confirm with the vendor which membrane and dosing-pump models are stocked against the in-country 9–12 month spares list.
How do I choose between suppliers for a Pyongyang project?
Shortlist vendors who can document the in-country spares path for RO membranes, control valves, and dosing pumps, and who will commit to a 9–12
